R Notebook

libraries

library(dplyr)
library(ggplot2)
library(nnet)
library(psych)
library(corrplot)
library(tidyverse)

package 㤼㸱tidyverse㤼㸲 was built under R version 3.5.3-- Attaching packages --------------------------------------- tidyverse 1.2.1 --
v tibble  2.1.3     v purrr   0.3.2
v tidyr   0.8.3     v stringr 1.4.0
v readr   1.3.1     v forcats 0.4.0
package 㤼㸱tibble㤼㸲 was built under R version 3.5.3package 㤼㸱tidyr㤼㸲 was built under R version 3.5.3package 㤼㸱readr㤼㸲 was built under R version 3.5.3package 㤼㸱purrr㤼㸲 was built under R version 3.5.3package 㤼㸱stringr㤼㸲 was built under R version 3.5.3package 㤼㸱forcats㤼㸲 was built under R version 3.5.3-- Conflicts ------------------------------------------ tidyverse_conflicts() --
x psych::%+%()        masks ggplot2::%+%()
x purrr::accumulate() masks foreach::accumulate()
x psych::alpha()      masks ggplot2::alpha()
x dplyr::between()    masks data.table::between()
x dplyr::combine()    masks randomForest::combine()
x tidyr::expand()     masks Matrix::expand()
x dplyr::filter()     masks stats::filter()
x dplyr::first()      masks data.table::first()
x dplyr::lag()        masks stats::lag()
x dplyr::last()       masks data.table::last()
x purrr::lift()       masks caret::lift()
x ggplot2::margin()   masks randomForest::margin()
x dplyr::slice()      masks xgboost::slice()
x purrr::transpose()  masks data.table::transpose()
x purrr::when()       masks foreach::when()

Load data

data<- read.csv("C://Users/s137r4//Desktop//Analysisfile//Accident14_17.csv")

Warning: closing unused connection 16 (<-CHR5C99E.corp.gwpnet.com:11346)
Warning: closing unused connection 15 (<-CHR5C99E.corp.gwpnet.com:11346)
Warning: closing unused connection 14 (<-CHR5C99E.corp.gwpnet.com:11346)
Warning: closing unused connection 9 (<-CHR5C99E.corp.gwpnet.com:11346)
Warning: closing unused connection 8 (<-CHR5C99E.corp.gwpnet.com:11346)
Warning: closing unused connection 7 (<-CHR5C99E.corp.gwpnet.com:11346)
Warning: closing unused connection 6 (<-CHR5C99E.corp.gwpnet.com:11346)

data <- data[, -c(1)]

missing values

p<- function(x){sum(is.na(x))/length(x)*100}
apply(data, 2, p)

                                          X                              Accident_Index                       Location_Easting_OSGR 
                                0.000000000                                 0.000000000                                 0.010534305 
                     Location_Northing_OSGR                                   Longitude                                    Latitude 
                                0.010534305                                 0.006224816                                 0.006224816 
                               Police_Force                           Accident_Severity                          Number_of_Vehicles 
                                0.000000000                                 0.000000000                                 0.000000000 
                       Number_of_Casualties                                        Date                                 Day_of_Week 
                                0.000000000                                 0.000000000                                 0.000000000 
                                       Time                  Local_Authority_.District.                   Local_Authority_.Highway. 
                                0.000000000                                 0.000000000                                 0.000000000 
                            X1st_Road_Class                            X1st_Road_Number                                   Road_Type 
                                0.000000000                                 0.000000000                                 0.000000000 
                                Speed_limit                             Junction_Detail                            Junction_Control 
                                0.000000000                                 0.000000000                                 0.000000000 
                            X2nd_Road_Class                            X2nd_Road_Number           Pedestrian_Crossing.Human_Control 
                                0.000000000                                 0.000000000                                 0.000000000 
    Pedestrian_Crossing.Physical_Facilities                            Light_Conditions                          Weather_Conditions 
                                0.000000000                                 0.000000000                                 0.000000000 
                    Road_Surface_Conditions                  Special_Conditions_at_Site                         Carriageway_Hazards 
                                0.000000000                                 0.000000000                                 0.000000000 
                        Urban_or_Rural_Area Did_Police_Officer_Attend_Scene_of_Accident                   LSOA_of_Accident_Location 
                                0.000000000                                 0.000000000                                 0.000000000 
                          Vehicle_Reference                                Vehicle_Type                     Towing_and_Articulation 
                                0.000000000                                 0.000000000                                 0.000000000 
                          Vehicle_Manoeuvre            Vehicle_Location.Restricted_Lane                           Junction_Location 
                                0.000000000                                 0.000000000                                 0.000000000 
                   Skidding_and_Overturning                   Hit_Object_in_Carriageway                 Vehicle_Leaving_Carriageway 
                                0.000000000                                 0.000000000                                 0.000000000 
                 Hit_Object_off_Carriageway                        X1st_Point_of_Impact                Was_Vehicle_Left_Hand_Drive. 
                                0.000000000                                 0.000000000                                 0.000000000 
                  Journey_Purpose_of_Driver                               Sex_of_Driver                               Age_of_Driver 
                                0.000000000                                 0.000000000                                 0.000000000 
                         Age_Band_of_Driver                        Engine_Capacity_.CC.                             Propulsion_Code 
                                0.000000000                                 0.000000000                                 0.000000000 
                             Age_of_Vehicle                           Driver_IMD_Decile                       Driver_Home_Area_Type 
                                0.000000000                                 0.000000000                                 0.000000000 
                         Casualty_Reference                              Casualty_Class                             Sex_of_Casualty 
                                0.000000000                                 0.000000000                                 0.000000000 
                            Age_of_Casualty                        Age_Band_of_Casualty                           Casualty_Severity 
                                0.000000000                                 0.000000000                                 0.000000000 
                        Pedestrian_Location                         Pedestrian_Movement                               Car_Passenger 
                                0.000000000                                 0.000000000                                 0.000000000 
                     Bus_or_Coach_Passenger          Pedestrian_Road_Maintenance_Worker                               Casualty_Type 
                                0.000000000                                 0.000000000                                 0.000000000 
                    Casualty_Home_Area_Type                                        Year 
                                0.000000000                                 0.000000000 

colz <- c(7,11,15:31,34:48, 50:67)
data[, colz]<-lapply(data[, colz], factor)

by_year_count <- data %>% select(Accident_Index, Year) %>% group_by(Year) %>% summarise(total.count=n())%>% mutate(total.count = total.count/sum(total.count)*100)
by_year_count

chart_year <- ggplot(data=by_year_count, aes(x=Year, y=total.count, fill=Year)) + geom_bar(stat="identity")+geom_text(aes(label=paste0(round(total.count, 2),"%")),position=position_dodge(width=0.9), vjust=-0.25)
chart_year

by_yearsev <- data %>% select(Accident_Index, Year, Accident_Severity) %>% group_by(Year, Accident_Severity) %>% summarise(total.count=n())%>% mutate(total.count = total.count/sum(total.count)*100)
by_yearsev

chart_yearsev <- ggplot(data=by_yearsev, aes(x=Year, y=total.count, fill=Accident_Severity)) + geom_bar(stat="identity")+geom_text(aes(label=paste0(round(total.count, 2),"%")),position=position_stack(vjust=0.25))
chart_yearsev

recoding

data <- data %>%
  mutate(Vehicle_Manoeuvre_category=case_when(Vehicle_Manoeuvre %in% 1:5~"Low_speed_manoeuvre",
                          Vehicle_Manoeuvre %in% 6:10 ~ "Turning_Manoeuvre",
                          Vehicle_Manoeuvre %in% 11:15 ~ "Lane_change_Manoeuvre",
                          Vehicle_Manoeuvre %in% 16:18 ~ "Going_ahead_Manoeuvre"
                        ))

manaouvre<- data %>% select(Accident_Index, Vehicle_Manoeuvre_category)%>% group_by(Vehicle_Manoeuvre_category) %>% summarise(t_count = n())%>% drop_na(Vehicle_Manoeuvre_category)%>% mutate(t_count = t_count/sum(t_count)*100)
manaouvre

manuevsev<- data %>% select(Accident_Index,Accident_Severity, Vehicle_Manoeuvre_category)%>% group_by(Vehicle_Manoeuvre_category, Accident_Severity) %>% summarise(t_count = n())%>% drop_na(Vehicle_Manoeuvre_category)%>% mutate(t_count = t_count/sum(t_count)*100)
manuevsev

data<-data%>%
  filter(Urban_or_Rural_Area %in% 1:2)
data<-data%>%
  filter(Light_Conditions %in% 1:7)
#1-highway 2- other roads 
data <- data %>%
  mutate(Highway=case_when( X1st_Road_Class %in% 1:2~ 1,
                            X1st_Road_Class %in% 3:6 ~ 0))
# 1 - daylight and 2- darkness 
data <- data %>%
  mutate(Light_Cond=case_when(Light_Conditions ==1 ~ 1,
                              Light_Conditions %in% 2:7 ~ 0))

light<- data %>% select(Accident_Index, Light_Cond)%>% group_by(Light_Cond) %>% summarise(t_count = n())%>% mutate(t_count = t_count/sum(t_count)*100)
light

light1<- data %>% select(Accident_Index, Light_Cond,Accident_Severity)%>% group_by(Light_Cond,Accident_Severity) %>% summarise(t_count = n())%>% mutate(t_count = t_count/sum(t_count)*100)
light1

table(data$Light_Cond)

     0      1 
121617 296039 

table(data$Highway)

     0      1 
394659  22997 

data <- data %>%
  mutate(Weather_Cond=case_when(Weather_Conditions==1 ~1,
                                Weather_Conditions %in% 2:8 ~0
                                      ))

weather<- data %>% select(Accident_Index, Weather_Cond)%>% group_by(Weather_Cond) %>% summarise(t_count = n())%>% drop_na(Weather_Cond)%>%  mutate(t_count = t_count/sum(t_count)*100)
weather

weather2<- data %>% select(Accident_Index, Weather_Cond,Accident_Severity)%>% group_by(Weather_Cond,Accident_Severity) %>% summarise(t_count = n())%>% drop_na(Weather_Cond) %>% mutate(t_count = t_count/sum(t_count)*100)
weather2

table(data$Highway)

     0      1 
394659  22997 

data <- data %>%
  mutate(Road_Surface_Cond=case_when(Road_Surface_Conditions==1 ~1,
                                      Road_Surface_Conditions %in% 2:7 ~0
                                     ))
#1- dry and 2- wet
table(data$Road_Surface_Cond)

     0      1 
129537 286113 

Road_Sc<- data %>% select(Accident_Index,Road_Surface_Cond)%>% group_by(Road_Surface_Cond)%>% drop_na(Road_Surface_Cond) %>% summarise(t_count = n())%>%  mutate(t_count = t_count/sum(t_count)*100)
Road_Sc

Road_Sc1<- data %>% select(Accident_Index,Road_Surface_Cond,Accident_Severity)%>% group_by(Road_Surface_Cond,Accident_Severity)%>% drop_na(Road_Surface_Cond) %>% summarise(t_count = n())%>%  mutate(t_count = t_count/sum(t_count)*100)
Road_Sc1

data <- data %>%
  mutate(Area=case_when(Urban_or_Rural_Area==1& Highway==1 ~ "Highway",
                        Urban_or_Rural_Area==2& Highway==1 ~"Highway",
                        Urban_or_Rural_Area==1& Highway==0 ~ "Urban",
                        Urban_or_Rural_Area==2& Highway==0 ~ "Rural"))
table(data$Area)

Highway   Rural   Urban 
  22997  149779  244880 

areads<- data %>% select(Accident_Index,Area)%>% group_by(Area) %>% summarise(t_count = n())%>%  mutate(t_count = t_count/sum(t_count)*100)
areads

areads1<- data %>% select(Accident_Index,Area,Accident_Severity)%>% group_by(Area,Accident_Severity) %>% summarise(t_count = n())%>%  mutate(t_count = t_count/sum(t_count)*100)
areads1

data <- data %>%
  mutate(spd_lt=case_when(Speed_limit %in% 1:20 ~"0-30",
                          Speed_limit %in% 20:30 ~ "30-60",
                          Speed_limit %in% 40:60 ~"60-100",
                          Speed_limit == 70 ~"100-130"))
table(data$spd_lt)

   0-30 100-130   30-60  60-100 
 113875   28838  178895   96048 

spdlt<- data %>% select(Accident_Index,spd_lt)%>% group_by(spd_lt) %>% summarise(t_count = n())%>%  mutate(t_count = t_count/sum(t_count)*100)
spdlt

spdlt1<- data %>% select(Accident_Index,spd_lt,Accident_Severity)%>% group_by(spd_lt,Accident_Severity) %>% summarise(t_count = n())%>%  mutate(t_count = t_count/sum(t_count)*100)
spdlt1

data <- data %>%
  mutate(Hit_object_on_carraige_way =case_when(Hit_Object_in_Carriageway == 1 ~ 1,
                          Hit_Object_in_Carriageway == 2 ~ 0,
                          Hit_Object_in_Carriageway == 4 ~ 1,
                          Hit_Object_in_Carriageway == 5:7 ~ 0,
                          Hit_Object_in_Carriageway == 9:12 ~ 0,
                          Hit_Object_in_Carriageway == 8 ~ 1,
                          Hit_Object_in_Carriageway == 0 ~ 0,
                          Hit_Object_in_Carriageway == 2 ~ 0,))

table(data$Hit_object_on_carraige_way)

obstacle<- data %>% select(Accident_Index,Hit_object_on_carraige_way)%>% group_by(Hit_object_on_carraige_way) %>% drop_na(Hit_object_on_carraige_way)%>% summarise(t_count = n())%>%  mutate(t_count = t_count/sum(t_count)*100)
obstacle

obstacle1<- data %>% select(Accident_Index,Hit_object_on_carraige_way,Accident_Severity,)%>% group_by(Hit_object_on_carraige_way,Accident_Severity) %>% drop_na(Hit_object_on_carraige_way)%>% summarise(t_count = n())%>%  mutate(t_count = t_count/sum(t_count)*100)
obstacle1

data<-data%>%
  filter(X1st_Point_of_Impact %in% 0:4)

table(data$X1st_Point_of_Impact)

impact<- data %>% select(Accident_Index,X1st_Point_of_Impact)%>% group_by(X1st_Point_of_Impact) %>% drop_na(X1st_Point_of_Impact)%>% summarise(t_count = n())%>%  mutate(t_count = t_count/sum(t_count)*100)
impact

impact1<- data %>% select(Accident_Index,X1st_Point_of_Impact,Accident_Severity)%>% group_by(X1st_Point_of_Impact,Accident_Severity) %>% drop_na(X1st_Point_of_Impact)%>% summarise(t_count = n())%>%  mutate(t_count = t_count/sum(t_count)*100)
impact1

pedestrian movement

data <- data %>%
  mutate(Pedestrian_Mvt=case_when(Pedestrian_Movement %in%  1:4 ~1,
                                  Pedestrian_Movement %in%  5:9 ~0,
                                  Pedestrian_Movement ==  0 ~0
                                      ))
table(data$Pedestrian_Mvt)

pedmove<- data %>% select(Accident_Index,Pedestrian_Mvt)%>% group_by(Pedestrian_Mvt)%>%  drop_na(Pedestrian_Mvt)%>% summarise(t_count = n())%>%  mutate(t_count = t_count/sum(t_count)*100)
pedmove

pedmove1<- data %>% select(Accident_Index,Pedestrian_Mvt,Accident_Severity)%>% group_by(Pedestrian_Mvt,Accident_Severity)%>%  drop_na(Pedestrian_Mvt)%>% summarise(t_count = n())%>%  mutate(t_count = t_count/sum(t_count)*100)
pedmove1

Casualty type

table(data$Casualty_Type)

casualty class this tow are interchangable and casaulty type

castyp<- data %>% select(Accident_Index,Casualty_Class)%>% group_by(Casualty_Class)%>% summarise(t_count = n())%>%  mutate(t_count = t_count/sum(t_count)*100)
castyp

castyp1<- data %>% select(Accident_Index,Casualty_Class,Accident_Severity)%>% group_by(Casualty_Class,Accident_Severity)%>% summarise(t_count = n())%>%  mutate(t_count = t_count/sum(t_count)*100)
castyp1

by variate analysis

highway_urban <- data %>% select(Accident_Index, Urban_or_Rural_Area, Highway )%>% group_by(Urban_or_Rural_Area, Highway ) %>% summarise(t_count = n())
highway_urban

chart_highway_urban<- ggplot(data=highway_urban, aes(x=Highway, y=t_count, fill=Urban_or_Rural_Area))+geom_bar(stat = "identity")+geom_text(aes(label=paste0(round(t_count, 2))), position = position_stack(vjust=0.9))+labs(y="Number of accidents", x= "Road type")+ scale_fill_discrete(name= "Area", labels=c("Urban", "Rural"))+
  theme(legend.position="bottom", legend.direction="horizontal",
        legend.title = element_blank())+ scale_fill_manual(values =c("#00BFFF", "#1E90FF"))

Scale for 'fill' is already present. Adding another scale for 'fill', which will replace the existing scale.

chart_highway_urban

area

areasev<- data %>% select(Accident_Index,Accident_Severity, Area)%>% group_by(Accident_Severity,Area) %>% summarise(t_count = n())%>% mutate(t_count = t_count/sum(t_count)*100)
areasev

chart_areasev<- ggplot(data=areasev, aes(x=Accident_Severity, y=t_count, fill= Area))+geom_bar(stat = "identity")+geom_text(aes(label=paste0(round(t_count, 2), "%")), position = position_stack(vjust=0.9))+labs(y="Percentage", x= "Accident severity")+ scale_fill_discrete(name= "Area", labels=c("Higheway", "Urban", "Rural"))+
  theme(legend.position="bottom", legend.direction="horizontal",
        legend.title = element_blank())#+scale_fill_brewer(palette = "Set2") #scale_fill_manual(values =c("#00BFFF", "#1E90FF", "#4169E1"))
chart_areasev

manuevsev<- data %>% select(Accident_Index,Accident_Severity, Vehicle_Manoeuvre_category)%>% group_by(Vehicle_Manoeuvre_category, Accident_Severity) %>% summarise(t_count = n())%>% drop_na(Vehicle_Manoeuvre_category)%>% mutate(t_count = t_count/sum(t_count)*100)
manuevsev

ADAS classification

data <- data %>%
  mutate(accident_type=case_when(Vehicle_Manoeuvre_category=="Going_ahead_Manoeuvre"& X1st_Point_of_Impact==1 & Number_of_Vehicles==2 & Junction_Detail==0 ~ "Rear-end collision",
                        Vehicle_Manoeuvre_category=="Turning_Manoeuvre" & X1st_Point_of_Impact %in% 3:4 & Number_of_Vehicles==2 &Junction_Detail %in% 1:9 ~ "Turning collision",
                        Vehicle_Manoeuvre_category=="Going_ahead_Manoeuvre" & Number_of_Vehicles==2 & X1st_Point_of_Impact %in% 2:3 &Junction_Detail %in% 1:9 ~ "Entering/crossing collision",
                        Vehicle_Manoeuvre_category=="Lane_change_Manoeuvre" & Hit_object_on_carraige_way==1 ~ "Collision with stationary traffic",
                        Vehicle_Manoeuvre_category=="Lane_change_Manoeuvre"& X1st_Point_of_Impact %in% 3:4 ~"collision with longitunal traffic",
                        Pedestrian_Mvt ==1 & Casualty_Class==3 ~ "Accident with pedestrian crossing the road",
                        Hit_object_on_carraige_way==1 & X1st_Point_of_Impact ==1 ~"Accident with other obsctacles on the carriageway"
                        ))

table(data$accident_type)

Accident with other obsctacles on the carriageway        Accident with pedestrian crossing the road                 collision with longitunal traffic 
                                             3547                                             50843                                              5272 
                Collision with stationary traffic                       Entering/crossing collision                                Rear-end collision 
                                              596                                             17666                                             31601 
                                Turning collision 
                                            15309 

acctype<- data %>% select(Accident_Index,accident_type  )%>% group_by(accident_type) %>% summarise(t_count = n())%>%drop_na(accident_type)%>% mutate(t_count = t_count/sum(t_count)*100)
acctype

chart_acctype<- ggplot(data=acctype, aes(x=accident_type , y=t_count, fill= accident_type ))+geom_bar(stat = "identity")+geom_text(aes(label=paste0(round(t_count, 2), "%")),position=position_dodge(width=0.9), vjust=-0.25)
chart_acctype

acctypearea<- data %>% select(Accident_Index,Area,accident_type)%>% group_by(accident_type,Area) %>% drop_na(accident_type) %>% summarise(t_count = n())%>% mutate(t_count = t_count/sum(t_count)*100)
acctypearea

chart_acctypearea<- ggplot(data=acctypearea, aes(x=accident_type, y=t_count, fill= Area ))+geom_bar(stat = "identity")+geom_text(aes(label=paste0(round(t_count, 2), "%")),position=position_stack(vjust=0.9))+labs(y="Percent %", x= "Accident Type")+ scale_fill_discrete(name= "Area", labels=c("Highway", "Urban", "Rural"))
chart_acctypearea

acctypspeed<- data %>% select(Accident_Index,spd_lt,accident_type)%>% group_by(accident_type,spd_lt) %>% drop_na(accident_type) %>% summarise(t_count = n())%>% mutate(t_count = t_count/sum(t_count)*100)
acctypspeed

chart_acctypspeed<- ggplot(data=acctypspeed, aes(x=accident_type, y=t_count, fill= spd_lt ))+geom_bar(stat = "identity")+geom_text(aes(label=paste0(round(t_count, 2), "%")),position=position_stack(vjust=0.9))+labs(y="Percent %", x= "Accident Type") #scale_fill_discrete(name= "Area", labels=c("Highway", "Urban", "Rural"))
chart_acctypspeed

severity, just slight and fatal as ADAS can not do much about slight

acctypsev<- data %>% select(Accident_Index,Accident_Severity,accident_type)%>% group_by(accident_type,Accident_Severity) %>% drop_na(accident_type) %>% summarise(t_count = n())%>% mutate(t_count = t_count/sum(t_count)*100)
acctypsev

chart_acctypsev<- ggplot(data=acctypsev, aes(x=accident_type, y=t_count, fill=Accident_Severity ))+geom_bar(stat = "identity")+geom_text(aes(label=paste0(round(t_count, 2), "%")),position=position_stack(vjust=0.9))+labs(y="Percent %", x= "Accident Type") #scale_fill_discrete(name= "Area", labels=c("Highway", "Urban", "Rural"))
chart_acctypsev

acctypall<- data %>% select(Accident_Index,Accident_Severity,accident_type, Area, spd_lt)%>% group_by(accident_type,Area, spd_lt,Accident_Severity) %>% drop_na(accident_type) %>% summarise(t_count = n())%>% mutate(t_count = t_count/sum(t_count)*100)
acctypall

dataa1 <- data
filter(dataa1, Age_of_Vehicle < 5 & Accident_Severity !=3) 

㤼㸱<㤼㸲 not meaningful for factors

NA

dataa1<-dataa1 %>%
  filter(Age_of_Vehicle %in% 1:6 & Accident_Severity %in% 1:2)

acctypall2<- dataa1 %>% select(Accident_Index,Accident_Severity,accident_type, Area, spd_lt)%>% group_by(accident_type,Area, spd_lt,Accident_Severity) %>% drop_na(accident_type) %>% summarise(t_count = n())%>% mutate(t_count = t_count/sum(t_count)*100)
acctypall2

acctype1<- dataa1 %>% select(Accident_Index,accident_type  )%>% group_by(accident_type) %>% summarise(t_count = n())%>%drop_na(accident_type)%>% mutate(t_count = t_count/sum(t_count)*100)
acctype

chart_acctype1<- ggplot(data=acctype1, aes(x=accident_type , y=t_count, fill= accident_type ))+geom_bar(stat = "identity")+geom_text(aes(label=paste0(round(t_count, 2), "%")),position=position_dodge(width=0.9), vjust=-0.25)
chart_acctype1

acctypspeed1<- dataa1 %>% select(Accident_Index,spd_lt,accident_type)%>% group_by(accident_type,spd_lt) %>% drop_na(accident_type) %>% summarise(t_count = n())%>% mutate(t_count = t_count/sum(t_count)*100)
acctypspeed1

chart_acctypspeed1<- ggplot(data=acctypspeed1, aes(x=accident_type, y=t_count, fill= spd_lt ))+geom_bar(stat = "identity")+geom_text(aes(label=paste0(round(t_count, 2), "%")),position=position_stack(vjust=0.9))+labs(y="Percent %", x= "Accident Type") #scale_fill_discrete(name= "Area", labels=c("Highway", "Urban", "Rural"))
chart_acctypspeed1

acctypsev1<- dataa1 %>% select(Accident_Index,Accident_Severity,accident_type)%>% group_by(accident_type,Accident_Severity) %>% drop_na(accident_type) %>% summarise(t_count = n())%>% mutate(t_count = t_count/sum(t_count)*100)
acctypsev

chart_acctypsev1<- ggplot(data=acctypsev1, aes(x=accident_type, y=t_count, fill=Accident_Severity ))+geom_bar(stat = "identity")+geom_text(aes(label=paste0(round(t_count, 2), "%")),position=position_stack(vjust=0.9))+labs(y="Percent %", x= "Accident Type") #scale_fill_discrete(name= "Area", labels=c("Highway", "Urban", "Rural"))
chart_acctypsev1

acctypearea1<- dataa1 %>% select(Accident_Index,Area,accident_type)%>% group_by(accident_type,Area) %>% drop_na(accident_type) %>% summarise(t_count = n())%>% mutate(t_count = t_count/sum(t_count)*100)
acctypearea1

chart_acctypearea1<- ggplot(data=acctypearea1, aes(x=accident_type, y=t_count, fill= Area ))+geom_bar(stat = "identity")+geom_text(aes(label=paste0(round(t_count, 2), "%")),position=position_stack(vjust=0.9))+labs(y="Percent %", x= "Accident Type")+ scale_fill_discrete(name= "Area", labels=c("Highway", "Urban", "Rural"))
chart_acctypearea1